A method for causal intervention on continuous variables shows verb bias is causally encoded in LLM steering vectors and affects syntactic preferences, though links to in-context learning error signals are not causal.
Understanding task vectors in in-context learning: Emergence, functionality, and limitations
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.CL 2years
2026 2verdicts
UNVERDICTED 2representative citing papers
A distributional alignment metric d_NTP and a linear regression method LTV for task vectors that improves accuracy by 9.2% over baselines on classification and regression tasks across multiple LLMs.
citing papers explorer
-
Causal Interventions on Continuous Variables: A Case Study on Verb Bias in Steering Vectors for In-Context Learning
A method for causal intervention on continuous variables shows verb bias is causally encoded in LLM steering vectors and affects syntactic preferences, though links to in-context learning error signals are not causal.
-
Distributional Alignment as a Criterion for Designing Task Vectors in In-Context Learning
A distributional alignment metric d_NTP and a linear regression method LTV for task vectors that improves accuracy by 9.2% over baselines on classification and regression tasks across multiple LLMs.